Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Intel might be too big to fail

tomshardware.com

Post reply on HN
Print view View original

141–150 of 253 posts

Re: Intel might be too big to fail

#141

Post by _heimdall »

exabrial wrote: post #114
_heimdall wrote: post #16

Earlier quoted context omitted.

I do believe that's where tariffs would come into play. They're a politically heated topic these days, but tariffs are meant to help manage exactly such a situation.

They also have been successful in pockets, but in my mind they're a blunt instrument and its rarely worked. And you're right about politically heated, thats probably the biggest problem. Not to sound like a Trumper, b/c I'm not: The media crucified Trump for installing a tariff on Chinese solar centers, which were being produced below cost with state sponsored material and labor. The Chinese intent was to keep Americ…

100% agree on both accounts. Tariffs often don't work, or only partly meet their original goals. They're especially bad if done as a blanket tariff rather than target on specific products or industries.

Not sure if Trump would actually be planning to add targeted tariffs or blanket tariffs. Hopefully not the later, if we're stuck with him rather than stuck with Harris.

Re: Intel might be too big to fail

#142

Post by spoaceman7777 »

People here seem to be filling in a lot of details from their imagination.

Intel's losses all revolve around upgrading and spinning up their new fabs, fabs for which there is no other global competition besides TSMC and Samsung. The latest process was delayed.

Once they start filling orders, they will be the ones filling the hole that could be left in US microchip capabilities that are in imminent danger, being largely based in Taiwan.

Their new desktop chips are cracked, their Gaudi 3 accelerators, that just started shipping five weeks ago, are a shade better than H100's, for half the cost, and with improved and much more affordable capabilities for interconnect, in addition to other features.

It's incredibly important for the US to continue its push to building up the domestic TSMC, Intel, and related chip infrastructure that it has so far funded (though, has largely not paid out, contributing even further to Intel's current situation).

Perhaps, if tariffs really do come in, we'll see the many fab-less chip designers that people are praising in this thread bring their manufacturing back to American manufacturers. As they damned well should.

Re: Intel might be too big to fail

#143

Post by matwood »

dahart wrote: post #124
vorpalhex wrote: post #56

Earlier quoted context omitted.

Every major communist country has been left in the dust by capitalist ones. Poverty is at a historical low. We have gone from people dying from starvation to complaining that not everyone can buy a house. Nearly everyone has a universally connected borderline supercomputer in their pocket.

Poverty has indeed been decreasing most places, but your framing is very misleading. Half of the people on earth still live on less than $7/day in 2017 PPP dollars, and they are not universally connected or supercomputing. Over 8% of the global population is still below the $2.15/day PPP threshold that defines “extreme poverty”, which means it’s not enough money to purchase enough food to survive, literally still dyi…

I'm not disagreeing with you that people are definitely struggling, but it's hard to see numbers like what you present w/o also numbers on what it cost to live per day. Is it an extreme case of CA vs WV or is it like someone making $7/day living in CA?

Re: Intel might be too big to fail

#145

Post by YZF »

toast0 wrote: post #122
insane_dreamer wrote: post #49

Earlier quoted context omitted.

Most companies should be banned from doing this.

Buybacks are just dividends with better tax implications that somehow make people angry. Companies historically are expected to pay dividends, at least when their business is doing well. Business at Intel was doing well for most of 1990-2017. There was some time after the Pentium 4 stopped scaling before the Pentium 4M offered a recovery, and the Itanium mess; but overall pretty good until 2017.

Buybacks aren't exactly like dividends because they directly affect the pricing of the stock by interfering with the supply and demand. That said I think people are mostly angry about the conflict of interest where CEOs that have current and future shares are making decisions by what will maximize their personal returns rather than what's best for the company and shareholders.

When a company with growth prospects does well it should invest those $$$'s into things like R&D and expansion. Companies that pay their profit as dividend are generally not expected to grow as much and their stock prices (P/E) tends to reflect that.

That said the taxation aspect is maybe a problem and should be addressed if it's not working as intended.

Re: Intel might be too big to fail

#146

Post by BeetleB »

jessfyi wrote: post #8

Any part of "saving" Intel should include a mechanism barring them from putting any more money that should be spent on R&D towards stock buybacks ($152B since 1990 as of September.) That said quoting the former Intel CEO (who still owns 3,245,986 shares) as "[one of the] expert[s] who says breaking up Intel won't do any good" seems like journalist malpractice--and makes me all the more certain it should be subsumed b…

Intel stopped buybacks a few years ago, and are now stopping dividends.

Re: Intel might be too big to fail

#147

Post by matwood »

toast0 wrote: post #122
insane_dreamer wrote: post #49

Earlier quoted context omitted.

Most companies should be banned from doing this.

Buybacks are just dividends with better tax implications that somehow make people angry. Companies historically are expected to pay dividends, at least when their business is doing well. Business at Intel was doing well for most of 1990-2017. There was some time after the Pentium 4 stopped scaling before the Pentium 4M offered a recovery, and the Itanium mess; but overall pretty good until 2017.

> Buybacks are just dividends with better tax implications that somehow make people angry.

Yeah, buybacks are the new boogieman. Let me decide when to take the tax hit on an investment and not be forced.

Re: Intel might be too big to fail

#148

Post by BeetleB »

BenFranklin100 wrote: post #144

Intel seems like it would benefit with an Elon Musk type of character in charge. Cut out the needless layers of management and go hard core engineering.

Unlike Twitter, fabs actually do need a large number of employees. And unlike whatever he's building at Tesla and SpaceX, changing a fab process is an order of magnitude more expensive, and the engineering/science is a lot more sophisticated. There's a reason semiconductor manufacturing is considered the most advanced manufacturing on the planet.

Re: Intel might be too big to fail

#149

Post by BeetleB »

mrweasel wrote: post #24

AMD and Intel just switched places, before Ryzen AMD was a failing company, threatening to not one give Intel a near monopoly, but also handing Nvidia a true monopoly. If AMD can be turned around, why can't Intel, a much larger company with more diverse product lines (not as many as previous, but still). It's not even that Intel aren't selling chips, I'd guess that they are still massively outselling AMD. The fact th…

> If AMD can be turned around, why can't Intel, a much larger company with more diverse product lines (not as many as previous, but still).

Because AMD is fabless and Intel isn't. The latter is spending huge amounts of money to increase their fab capacity, and have positioned themselves to succeed only if they can get external Foundry customers.

If Intel split into two companies, then yes, your point is valid for the fabless portion. But it would also mean the fab portion would die almost immediately (they currently are operating on a loss) - they need both the US bailout and the fabless revenue to survive.

Re: Intel might be too big to fail

#150

Post by mitjam »

Intel is a huge asset, they missed major industry changes, like EUV, ARM, GPU, and splitting design and foundry. I still believe they can come back strong, but need bold decisions to regain defensible competitive advantages.
Post reply on HN